Skip to main content

Dear All,

The attached vibration spectra from one of our chilled water pumps. I suspect this may be a cracked rotor bar. Pumps are usually changed over every month and I'm planning to do a single phase rotor test at the changeover next month. Based on the spectra would it be safe to have this pump running till then??

Any input will be much appreciated   

Attachments

Original Post

I suggest conducting motor current spectrum analysis (MCSA) to confirm it is/not an electrical rotor fault. My quick diagnosis is bearing fault (looseness), so that is why I suggested MCSA. I have seen vibration data indicating a rotor fault, as you suspect, but it proved to be a bearing fault. This was confirmed on motors with both rolling element and journal bearings! Confirming the problem before shutdown can be a very good strategy!

Walt

@Walt Strong posted:

I suggest conducting motor current spectrum analysis (MCSA) to confirm it is/not an electrical rotor fault. My quick diagnosis is bearing fault (looseness), so that is why I suggested MCSA. I have seen vibration data indicating a rotor fault, as you suspect, but it proved to be a bearing fault. This was confirmed on motors with both rolling element and journal bearings! Confirming the problem before shutdown can be a very good strategy!

Walt

Thanks Walt for the reply. Unfortunately MSCA is not an option at the moment as we're waiting on the specific current clamp for the analyser we're using and with the covid situation that might be a while. 

No intention of shutting down the pump just for the rotor test. We have 2 pumps we run in turn every other month so this one will be off service anyways. 

I've also attached the time signal and the envelope spectrum here with BPFI harmonics. No significant increases there over time. So I'm thinking if there no rotor bar issues this pump is ok to run for a while under close watch. I thought of the rotor test I guess just for peace of mind since the pump is shut down anyways.

Any thoughts? 

 

 

Attachments

One characteristic of broken or cracked rotor bars is an audible beat.  The beat will increase in amplitude as load is increased.  Typical in the spectrum will be a large 1X with sidebands spaced at slip frequency (and multiples).

With a broken rotor bar, the amplitude of the beat increases with load. A cracked rotor bar can also cause localized heating of the rotor which causes uneven expansion and rotor bowing. This results in an unbalance and a strong 1 times running speed vibration as well as the side bands related to slip frequency.

There is an in depth discussion at https://www.maintenance.org/to...-rotor-bars-symptoms and an old but still good paper at Trouble Shooting Induction Motors.

 

 

 

I measured several in a power plant near Philadelphia, but I did not notice an audible sound; plant noise may have made this difficult.

I did find one in a motor shop test of the motor alone, but I did not hear it either.  The side bands were present.

Well we can’t have that situation!  

Go to https://www.youtube.com/watch?v=7xGlz1MVHpk and have a listen.  The narration refers to the noise as β€œpulsating.”  You may remember JK from the Philly office (Broomall) and he documented a case where the audible beat matched a beat apparent in a waveform from a velocity coil. 

@Walt Strong posted:

Did BN ever develop a Prox. Probe Microphone?

Walt

Walt, at least up to the end of 2018, when I finally decided to 100% leave, BN had not done anything that was related to acoustics.  Certainly not within the ADRE product line but I can't speak for portables.

I was on the design team for the 408 and I pushed for the capability to attach multiple microphones for sound pressure and sound power measurements as are often required in FAT testing.  The 408 had the hardware capability and the software package would support the processing of the data.  I never was able to interest the really high echelon of management in the techniques.  I think in large part, even though BN had fantastic engineers, few had experience with any form of machinery acoustics.  

Thimba, are you getting an audible beat (I asked earlier)?

Hi John,

Thanks for your reply. 

I can't get to the plant at the moment with border closures due to covid (not sure when I could the way things are). I've asked our technician who does the measurements in my place and waiting for an update and audio recording and a long time waveform. I will post as soon as I receive info. 

 

 

 

John stated "One characteristic of broken or cracked rotor bars is an audible beat.  The beat will increase in amplitude as load is increased."

I agree and have witnessed this several times where the rotor had a confirmed electrical fault. I have also witnessed the same/similar sound when a bearing was very loose/worn, so it was a mechanical fault that was excited by electrical source.

It is a worthwhile to have both vibration and motor current/flux data to resolve the fault diagnosis before scheduling repairs and motor teardown.

Walt

John,

"I think in large part, even though BN had fantastic engineers, few had experience with any form of machinery acoustics."

I think it is worthwhile for anyone doing machine condition monitoring to have a basic understanding of acoustics and be able to make sound pressure measurements. This is intended to be a general comment. I have been drawn into many machine diagnostic situations because the machine or component had perceived high or "unusual" sound/noise. I carry a small sound meter with AC signal output in my portable analyzer case and have several other types of microphones. I have a dual background in vibrations and acoustics and noise control.

Walt

Thimba, are you getting an audible beat (I asked earlier)?

Hi John,

Apologies for the late response. I just got my update and additional measurements from the plant. (I still cant get to the plant due to covid restrictions)

From the audio I received, I cant really pick an audible beat. However there's a fair bit of noise from nearby pumps so its near impossible to pick a beat even if there was one.  

The TWF does seem to show beats at ~2.1Hz

I've also attached high res velocity spectra and High frequency spectrum. 

The pass-band for the envelope spectrum @Walt Strong mentioned showing BPFI harmonics is set to 7-12 kHz. Based on the high frequency spectrum, I'm thinking if this is a bearing defect, it'll be in early stages and happy run this and monitor closely. I'm hoping to do the single phase rotor test early next month. I will post what we find there. 

Thank you both for your thoughts on this topic. 

Thimba 

 

 

 

 

Attachments

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×